GIS with R: slides for the unconverted

Francisco Rodriguez-Sanchez
@frod_san

2015-11-14

R: not only for stats

R can make beautiful maps

Made in R

http://spatial.ly/2012/02/great-maps-ggplot2/

Made in R

http://paulbutler.org/archives/visualizing-facebook-friends/

Made in R

http://oscarperpinan.github.io/spacetime-vis/

Made in R

http://oscarperpinan.github.io/spacetime-vis/

Made in R

http://oscarperpinan.github.io/spacetime-vis/

And it's very easy!

Map species occurrences in 2 lines of code

plot(gmap(occdata, type = "satellite"))
points(occdata, col = "red", pch=20, cex = 2)

plot of chunk unnamed-chunk-2

Let's go step by step

A dataframe of species occurrences

species lon lat
1 Solanum acaule Bitter -66.10 -21.90
2 Solanum acaule Bitter -71.00 -13.50
52 Solanum acaule Bitter -66.43 -24.22
53 Solanum acaule Bitter -72.07 -13.35
54 Solanum acaule Bitter -68.97 -15.23
55 Solanum acaule Bitter -64.95 -17.75

Make it a spatial object

coordinates(occs) <- c("lon", "lat")
str(occs, 2)

Formal class 'SpatialPointsDataFrame' [package "sp"] with 5 slots
  ..@ data       :'data.frame': 49 obs. of  1 variable:
  ..@ coords.nrs : int [1:2] 2 3
  ..@ coords     : num [1:49, 1:2] -66.1 -71 -66.4 -72.1 -69 ...
  .. ..- attr(*, "dimnames")=List of 2
  ..@ bbox       : num [1:2, 1:2] -72.5 -24.2 -64.7 -12.5
  .. ..- attr(*, "dimnames")=List of 2
  ..@ proj4string:Formal class 'CRS' [package "sp"] with 1 slot

Specify projection (CRS)

crs(occs) <- CRS("+proj=longlat +ellps=WGS84 +datum=WGS84")

See http://spatialreference.org

Project to Mercator and plot

plot(gmap(occs, type = "satellite"))
points(Mercator(occs), col = "red", pch = 20, cex = 2)

plot of chunk unnamed-chunk-6

Alternatively, load shapefile

occs <- shapefile("occs.shp")
plot(gmap(occs, type = "satellite"))
points(Mercator(occs), col = "red", pch = 20, cex = 2)

plot of chunk unnamed-chunk-8

Using ggmap

library(ggmap)
map <- get_map(bbox(occs), maptype = "watercolor", source = "stamen")
ggmap(map) +
  geom_point(aes(x = coords.x1, y = coords.x2), data = as.data.frame(coordinates(occs)),
             colour = "red", size = 4)

plot of chunk unnamed-chunk-9

Raster data

Download elevation data

elevation <- getData("alt", country = "ESP")

library(rasterVis)
levelplot(elevation)

plot of chunk unnamed-chunk-11

Dynamic interactive maps with leaflet

library(mapview)
mapView(occs)

Remote sensing growing fast

e.g. RStoolbox

Doing GIS in R: main advantages

  • Fully-reproducible scripts

  • Harness all R stats power

    • Data wrangling
    • Modelling
    • Dataviz

  • Easy!

Calling GIS from R

Includes calling R from ArcGIS too

To read more

http://pakillo.github.io/R-GIS-tutorial